Learning interpretable continuous-time models of latent stochastic dynamical systems.
Lea DunckerGergo BohnerJulien BoussardManeesh SahaniPublished in: ICML (2019)
Keyphrases
- dynamical systems
- predictive state representations
- latent variable models
- differential equations
- reinforcement learning methods
- dynamic systems
- connectionist networks
- dynamical behavior
- probabilistic model
- prior knowledge
- linear dynamical systems
- latent variables
- partially observable
- stochastic models
- linear systems
- dynamical models
- state space
- ordinary differential equations
- markov processes
- control theory
- learning algorithm
- hidden variables
- learning problems
- complex systems
- bayesian networks